Long-acting pre-exposure prophylaxis (LAP) has the potential to change HIV prevention across sub-Saharan Africa, but its public-health impact and value-for-money depend on programme ability to reach people at greatest risk. The authors modelled the effects of different allocation rules for LAP to quantify how targeted deployment could alter infections averted and programme costs up to 2050. The analysis focuses on adults aged 15 years or older in 34 countries that together represent 98% of adults living with HIV in the region.
A deterministic, age-structured compartmental model of HIV transmission among adults was developed for the included countries. Key model assumptions reported in the source include an assumed scale-up of ART to 95% coverage in every country by 2030 and an effectiveness of LAP to reduce HIV acquisition of 96%. The model evaluated allocation of a fixed quantity of LAP under alternate targeting rules and examined resulting epidemiological and economic outcomes through 2050. The manuscript and supplementary material are reported as containing the full data and methods.
Allocation strategies examined in the model ranged from uniform distribution of LAP across populations to allocation by routinely available demographic signals (district, sex and age) and an upper-bound “perfect” targeting scenario that assumes ideal identification of individuals at highest risk within demographic groups.
When a fixed supply of LAP was allocated by district, sex and age, the number of infections averted was between 2 and 21 times greater than with uniform allocation across the 34 countries. The perfect within-group risk identification scenario further increased impact to 5–32 times that of uniform allocation. These multipliers varied by country-level epidemic context, with the largest relative gains from targeting observed in lower-prevalence countries where absolute numbers of infections averted were smaller.
The effect of targeting was heterogeneous by epidemic burden. In higher-burden settings the reported multipliers (relative to uniform allocation) were smaller: 2–6 for allocation by district, sex and age, and 5–11 for perfect targeting, reflecting larger absolute impact under uniform strategies in such settings. Conversely, low-prevalence countries showed larger multipliers, meaning that targeting substantially increased efficiency although absolute infections averted remained smaller than in high-burden settings.
The authors modelled a scenario in which 3 million people use LAP starting in 2028 in the 11 countries then receiving Global Fund-supported lenacapavir. Under this scenario, the estimated infections averted by 2050 were:
These estimates illustrate how targeting rules influence both absolute and relative impacts when supply is constrained.
The modelling included discounted future ART costs averted by preventing infections and compared these savings against LAP programme costs. The analysis reports break-even thresholds for the fully loaded cost per person-year of LAP such that discounted averted ART costs fully offset LAP costs by 2050. The reported thresholds were approximately US$50 per person-year with perfect risk targeting, US$30 per person-year with district/sex/age targeting, and US$2 per person-year with uniform allocation. These values reflect the stronger economic case for LAP when it is effectively targeted to people at highest risk.
A scenario expanding LAP to 15 million users by 2035 was modelled across sub-Saharan Africa. Under the assumption that ART achieves 95% coverage by 2030 in every country, the model projects that adult HIV incidence across the region would fall below 0.1 per 1000 person-years by 2049 — approximately a decade earlier than the projection without LAP. This projection underscores the potential population-level effect of large-scale, targeted LAP deployment combined with high ART coverage.
The authors conclude that programme targeting is a key determinant of both epidemiological impact and net cost outcomes for LAP deployment. Allocation by district, sex and age is presented as a currently feasible approach that markedly increases infections averted per person-year of LAP relative to uniform distribution. The modelling suggests that LAP will offset its costs through averted ART only when programmes can target recipients and when the full programme cost per person-year is near the lower end of cost estimates.
The analysis also indicates that improved identification of individuals at highest risk within demographic groups, and sustained use among those individuals, could add substantially to impact. The authors propose that the same targeting principles could guide LAP allocation in other settings where HIV risk is concentrated. The source notes that full methodological detail, supplementary material and data are contained in the manuscript and its supplementary files.
Note: This work is reported as a preprint and has not been peer reviewed; it should not be used to guide clinical practice without corroborating evidence and peer review.